Interfractional patient setup uncertainty and anatomy change are widely recognized as one of the major limiting factors for maximum exploitation of modern radiation therapy techniques, such as intensity modulated radiation therapy (IMRT). Up to this point, almost all research efforts have been focused on reducing the adverse effects of organ movement/deformation by attempting to reposition the patient more accurately. Clinically, IMRT treatment plan optimization and dose delivery are still two decoupled steps, with the geometric uncertainties taking into account by population based margins encompassing the clinical target volume, which significantly compromises the success of radiation therapy. The recent advent of onboard volumetric imaging device provides a valuable tool for us to obtain 3D or even 4D geometric model of the patient in the treatment position and allows adaptive modification of IMRT plan during a course of treatment. The objective of this project is to develop enabling computational tools for image guided adaptive radiation therapy (IGART) and to show the potential clinical impact of the new paradigm of IGART. The underlying hypothesis of this work is that IGART will greatly reduce the uncertainty in beam targeting and provide substantially improved dose distributions required to achieve greater local tumor control while reducing the probability of normal tissue complications. Specific aims of the project are (1) to establish a cone beam CT (CBCT)-based dose reconstruction method for fractional and cumulative dose calculations; (2) to setup a dynamic closed-loop framework of IGART planning; and (3) to demonstrate the potential clinical impact of the proposed IGART. Execution of the project will demonstrate that the IGART is achievable and determine the level of improvement of IGART over the conventional IMRT. Given its significant promise in optimally compensating for interfractional geometric uncertainties as well as dosimetric errors incurred in previous fractions, successful completion of the project should lead to substantial improvement in cancer patient care. PUBLIC HEALTH RELEVANCE: Currently, a radiation therapy treatment plan is produced based on the patient's anatomical model from planning CT images acquired a few days or even weeks before treatment. Numerous investigations have revealed that there can be significant changes in the patient anatomy from day to day due to patient positioning uncertainties and physiologic and clinical factors. This project is aimed to develop enabling computational tools for a new paradigm of radiation therapy, referred to as image-guided adaptive radiation therapy (IGART), to eliminate the influence of inter-fractional anatomy change. IGART improves current radiation therapy by adaptively adjusting the beam parameters according to volumetric imaging data acquired with the patient in the actual treatment position.